Direct Access to Axially Substituted Subphthalocyanines from Trimethylsilyl-Protected Nucleophiles

(1)

Repositorio Institucional de la Universidad Autónoma de Madrid https://repositorio.uam.es

Esta es la versión de autor del artículo publicado en:

This is an author produced version of a paper published in:

Organic Letters 17.19 (2015): 4722-4725

DOI: http://dx.doi.org/10.1021/acs.orglett.5b02213

Copyright: © 2015 American Chemical Society.

El acceso a la versión del editor puede requerir la suscripción del recurso

Access to the published version may require subscription

(2)

1

Direct Access to Axially Substituted Subphthalocyanines from Trimethylsilyl-protected Nucleophiles

Julia Guilleme,

^†

Lara Martínez-Fernández,

^‡

Inés Corral,

^‡

Manuel Yáñez,

^‡

David González-Rodríguez

^†,

* and Tomas Torres

^†,§,

*

† Departamento de Química Orgánica, Facultad de Ciencias, Universidad Autónoma de Madrid, E-28049 Madrid, Spain.

‡ Departamento de Química, Facultad de Ciencias, Universidad Autónoma de Madrid, E-28049 Madrid, Spain.

§ IMDEA Nanociencia, c/Faraday 9, Campus de Cantoblanco, 28049, Spain.

Supporting Information Placeholder

ABSTRACT: A new synthetic one-step approach to perform the axial ligand exchange reaction in subphthalocyanines that employs trimethylsilyl-protected nucleophiles as starting materials is reported. Theoretical calculations indicate that the exchange reaction proceeds through a similar 4-centered σ-bond metathesis transition state as the substitution with phenols. This direct method allowed us to synthesize new axial derivatives of singular importance within the chemistry of subphthalocyanines, whose reactivity and X- ray crystalline structure was studied.

Subphthalocyanines (SubPcs) are unique nonplanar aromatic macrocycles composed of three diiminoisoindoline N-fused units arranged in a cone-shaped structure. These molecules are only known as boron complexes, whose tetrahedral coordination induces the deviation from planarity and thus, their concave nature.¹ Due to their extraordinary spectral and electronic fea- tures, SubPcs are considered competent candidates of increas- ing interest in different applied fields such as organic light-emit- ting diodes (OLEDs),² organic photovoltaic cells (OPVs),³ or non-linear optics.⁴ For these reasons, the design of simple and general synthetic routes that allow the incorporation of SubPcs into more complex systems still remains an active task.

The reactivity of SubPcs has been addressed by several au- thors over the last decades,^1,5 and can be classified in three different groups depending on the reactive center:¹ (i) axial reactivity (central boron atom), (ii) peripheral reactivity (functional groups attached to the aromatic carbons) and (iii) ring opening reaction (imine-type core of the SubPc). Among them, the axial ligand exchange is by far the most common approach employed to functionalize subazamacrocycles without altering the electronic properties of the macrocycle.^1c Nevertheless, the axial substitution in SubPcs is slower and tougher than the analogous process in subporphyrins, which usually are reported as meth- oxy-compounds.^1b,6

Recently, we reported a mechanistic study of the axial ligand exchange reaction between halo-SubPcs and phenols in which we determined a second-order reaction rate law for this process, both dependent on SubPc and phenol concentrations.⁷ We pro- posed that this transformation takes place through a -bond metathesis mechanism involving a bimolecular transition state in which the phenol assists in the dissociation of the polar boron- halogen bond by proton coordination, while the new B–O bond is being formed. These observations led us to conclude that in order to perform an axial ligand exchange reaction, a good nucleophile is not strictly required, but rather an agent that has some affinity for the axial halogen and hence can weaken the boron-halogen bond before actual substitution.

Encouraged by these results, we thought that using alcohols or other nucleophiles that are “protected” with a trialkylsilyl group (R3Si–Nu) may lead to a similar transition state, since silanes are well-known halophilic electrophiles.⁸ The initial chlorine atom of the SubPc would be now coordinated to the halophilic Si center in the transition state promoting the weakening of the B-Cl bond. Our assumption would be supported by previous work from Kato et al., who reported in 2006 the syn- thesis of a boron SubPc cation as a salt with a weakly coordinated carborane anion by reaction between Et3Si(CHB11Me5Br6) and the corresponding chloro-SubPc.^8c Moreover, the use of trimethylsilyl-protected reagents presents

(3)

2

several advantages. For instance, the byproduct (Me3SiX) that would be generated in this axial ligand exchange would be vol- atile in the reaction media, which should shift the equilibrium towards the products and simplify the purification step. In addition, their lower nucleophilicity in comparison with the depro- tected nucleophile analogues (Nu-H) should generate lower amounts of SubPc decomposition products. In this work, we re- port on the theoretical and experimental study of a new synthetic route to axially substituted SubPcs by direct reaction of chloro-SubPcs and trimethylsilyl (TMS)-protected nucleophiles (Scheme 1). We also evaluate the scope of this novel synthetic methodology and its potential to prepare novel SubPc de- rivatives that are otherwise not accessible by other methods.

N N N

N N

N B

Cl R²

R²

R² R² R² R²

Me3Si-Nu solvent (reflux)

Me₃SiCl

1 R1= -CH₃; R₂= H 2 R1= R₂= H 3 R1= R₂= F

R¹ R²

O

O Br

O

S O

O

O CF₃

N3

CN

CN 4 R1= -H; R₂= -H; Nu= 5 R₁= -CH₃; R₂= -H;Nu= 6 R1= -CH₃; R₂= -H;Nu= 7 R1= -F; R₂= -F; Nu= 8 R1= -H; R2= -H; Nu= 9 R₁= -H; R₂= -H; Nu= 10 R1= -CH₃; R₂= -H;Nu= R¹R²

R¹ R²

N N N

N N

N B

Nu R²

R²

R² R² R² R²

R¹ R²

R¹R²

R¹ R²

Scheme 1. Synthesis of axially substituted SubPcs by direct re- action between chloro-SubPcs and TMS-nucleophiles.

We initiated our studies with a classical, well-known reaction in SubPc chemistry: the axial ligand exchange with phenol derivatives. This reaction, as mentioned before, has been thor- oughly studied in our laboratories, both from the experimental and theoretical perspective.⁷ As a proof of concept and to make a direct comparison with our mechanistic study, we reacted chloro-SubPc 2 with O-TMS-phenol or phenol in exactly the same conditions in refluxing toluene. The desired phenoxy- SubPc 4 was obtained in 89% (entry 1, Table 1) using PhOTMS or 85% employing PhOH. This transformation took place in a few hours using an excess of the TMS derivative and no decomposition products were observed by TLC. The reaction with PhOH progressed initially faster but it took longer to reach com- pletion. We attribute this effect to the presence of remaining HCl in the reaction media at high conversion, which can shift the equilibrium towards the starting chloro-SubPc reagent. As expected, other substituted TMS-phenols were reactive as well.

For instance, the axial ligand exchange with the (4-bromophe- nol)trimethylsilane also proceeded successfully yielding SubPc 5 (entry 2, Table 1).

To investigate the reaction mechanism at a molecular level for trialkylsilyl protected nucleophiles we have mapped the ground state potential energy surface for SubPcCl 2 and the nu-

cleophile O-TMS-phenol by locating the corresponding station- ary points along the axial substitution reaction coordinate. Fig- ure 1 depicts the potential energy profile as predicted by density functional theory (DFT) calculations. Further computational details can be found in the Supporting Information. Similarly to what was found for unprotected phenol nucleophiles,⁷ the axial ligand exchange reaction is facilitated by the interaction of the two reagents (pre-reaction complex Reac2 in Figure 1) and pro- ceeds via a bimolecular transition state to a post-reaction com- plex (Prod2 in Figure 1), where the leaving Me3SiCl group and the phenoxy-SubPc are still interacting.

Table 1. TMS derivatives employed for the axial ligand ex- change reaction.

entry SubPc TMS-Nu yield/%

[a] product

1 2 89^[c] 4

2 1 66^[c] 5

3 1 - 6^[b]

4 1 - -

5 3 62 7

6 2 79^[c] 8

7 2 - -

8 2 - -

9 2 - -

10 2 -^[c] -

11 2 52^[d] 9

12 1 54^[d] 10

[a] Yields calculated with respect to the starting SubPc; [b] Traces of the product were detected; [c] In toluene; [d] In nitrobenzene.

From the geometric viewpoint, the formation of the pre-reaction complex Reac2 assists the nucleophilic attack by activating the B-Cl bond, which slightly stretches (0.01 Å), and by simul- taneously approaching the O atom of the phenol nucleophile to the B reaction centre. A further respective reinforcement and weakening of the B-O and B-Cl bonds is observed at the TS structure, concomitant to the imminent rupture and formation of the Si-O and Cl-Si bonds, before the new B-O bond consoli- dates in the post-reaction complex, where the phenoxySubPc is formed. Interestingly, the computed barrier is 17 kcal/mol larger than for the reaction with the unprotected nucleophile.

This can be attributed, among other reasons, to the large size of the silane moiety that hinders the preparation of nucleophilic attack in the pre-reaction complex and the arrangement of the incoming and leaving nucleophile at the 4-centre transition state. This increase in the energy barrier is consistent with the

(4)

slower reaction rates observed for the TMS-protected phenol in comparison with the unprotected phenol (see above).

Figure 1. Reaction profile for the axial ligand substitution reaction as predicted by DFT calculations. Energies in kcal/mol.

In order to get further insight into the reactivity of SubPc with protected nucleophiles, we next examined and compared the same reaction with aliphatic alcohols. In the case of the reaction with chloro-SubPcs, aliphatic alcohols are known to produce lower yields than aromatic alcohols mainly due to SubPc de- composition, presumably via ring-opening, after prolonged heating times. However, under the simple standard conditions employed in this work (see S.I.), that involve the use of only a small excess of the nucleophile, the reaction did not afford the corresponding axially substituted SubPcs after 1 day of reaction, either using the TMS-protected or the unprotected alcohol.

More drastic conditions and additional solvents (refluxing nitrobenzene or benzonitrile) were also tested but, unfortunately, only traces of SubPc product were detected in the reaction me- dia. The formation of the corresponding alkoxy-SubPc 6 could be favored in the presence of a much larger excess of nucleophile, but we did not analyze this situation because it is not practical from the synthetic point of view. On the other hand, the silyl enol ether (entry 4, Table 1), which can be viewed as a masked ketone, rapidly provoked the decomposition of the macrocycle.

We then selected different TMS derivatives other than alcohol nucleophiles, such as the triflate (Me3SiOTf) or azide (Me3SiN3). We published a procedure in the past with which the generation of triflate-SubPcs was carried out successfully.⁹ These interesting derivatives, however, were only produced in situ as transient activated intermediates for the axial substitution with other nucleophiles,⁹ but could not be isolated due to their rapid hydrolysis. A decrease in reactivity could be achieved by using electron-withdrawing macrocycles, such as 3. The result- ing perfluorinated-triflate-SubPc was much more inert and could be isolated by standard column chromatography in 62%

yield. In the case of Me3SiN3, the axial ligand exchange took place efficiently under the standard conditions (i.e, toluene at 110 °C) and the product, the first example of a SubPc with an axial B-N3 bond, was perfectly stable.

In order to explore the limits of this new synthetic route for axial ligand exchange in SubPcs, the reaction was next carried out with different TMS-carbon nucleophiles. We thought it

would be highly interesting to test Csp² (phenyl-) and Csp (al- kynyl- or nitrile) trimethylsilanes (entries 7-9, Table 1), which would generate novel SubPcs with B-C bonds. However, these reagents exhibited an evident lack of reactivity and only decomposition products were observed after days in refluxing toluene or nitrobenzene. The only exception was the Me3SiCN reagent, whose nucleophilicity seems to be the actual limit of this reaction for practical purposes. However, the reaction had to be carried out in the more polar nitrobenzene solvent in order to increase the reaction rate (see our previous work)⁷ and at temper- atures close to its boiling point. In this way, the valuable cyano- SubPcs 9 and 10 could be obtained in moderate yields.

On the other hand, the effect of the electronic nature of the macrocycle was also analysed. Both SubPcs functionalized with peripheral electron donating- (entries 2 and 12, Table 1) or electron-withdrawing groups (entry 5, Table 1) react with these TMS-nucleophiles. Their reactivity follows the same order as in any other axial-ligand exchange reaction: electron-rich Sub- Pcs react faster than electron-poor SubPcs. All new products were characterized by ¹H NMR, ¹³C NMR, UV-Vis, FT-IR spectroscopy, MS, and HRMS. Suitable crystals for X-ray dif- fraction analysis were obtained for 7, 8 and 9 and their resolved structure is shown in Figure 2.

Figure 2. Molecular structure of SubPcs 7 (a), 8 (b) and 9 (c) in the crystal as determined from X-ray diffraction analysis.

The new SubPcs 7, 8, 9 and 10 could be regarded as useful organic synthons to prepare a wide variety of other axially substituted products. Subsequently, we set out to investigate the be- haviour of these common organic functional groups when at- tached to the SubPc macrocycle. Triflate-SubPcs (7) have al- ready demonstrated to be extraordinary activated intermediates to efficiently substitute the halogen atom with diverse nucleophiles.⁹ On the other hand, we tested a set of common chemical transformations of nitriles (nucleophilic addition, hydrolysis and reduction reactions) or azides (reductions or 1,3-dipolar cy- cloadditions). These reactions were considered as highly valuable, since they could lead to a new family of “exotic” SubPcs bearing formyl, ketone or amine functional groups, among oth- ers, at the axial position of SubPcs. However and despite all our efforts, which are summarized in the S.I., these nitrile and azido groups displayed a rather unusual reactivity and none of these reactions worked in our hands. Either the axial functional group did not react under classical mild conditions or the SubPc macrocycle decomposed when using harsher conditions. All these

(5)

4

premises denote that these new SubPcs do not present the expected reactivity related with these common functional groups, probably as consequence of the electronic and/or steric effect exerted by the boron macrocycle.

The only exception was the 1,3-dipolar cycloaddition reac- tion to azide-SubPc 8 (Scheme 2). It is known that the chemistry and reactivity of boryl azides is rare and so it is their role as 1,3- dipoles in cycloaddition reactions.¹⁰ To the best of our knowledge only a few examples have been reported in which electron-rich boryl azides are able to thermally react with electron-poor alkynes, alkenes and nitriles to afford the corresponding cycloaddition adducts.¹⁰ SubPc 8 was subjected to a copper (I)-catalyzed cycloaddition (CuAAC) with phenylacetylene in THF and the corresponding triazole-SubPc 10 was obtained in a reasonable 30% yield, despite the absence of electron-withdrawing groups attached to the acetylene moiety. The cycload- duct was obtained as a single 1,4-regioisomer due to obvious steric hindrance. Although these conditions are far from the typ- ical “click” reactions, which usually proceed quantitatively in the presence of catalytic amounts of Cu⁺, this “click” approach may still be regarded as a convenient procedure to attach these functional macrocycles to more complex systems, like electro- active multi-component systems or polymers.

N N N

N N

N B

N₃

N N N

N N

N B

N

CuI, DIPEA, THF 50 ºC

N N

8 11

Scheme 2. Cycloaddition reaction between SubPc 8 and ethynylbenzene.

In conclusion, we have described a new synthetic route to carry out the axial ligand exchange reaction in SubPcs in one step that employs TMS-protected nucleophiles as starting materials. Theoretical calculations indicate that the ligand exchange reaction proceeds through a similar 4-centered -bond metathesis transition state as the substitution with phenols, where the TMS group concomitantly coordinates to the leaving chlorine atom and initiates the nucleophilic attack at the boron atom. Although this procedure cannot be established as a general and practical methodology, it can be considered as a good alternative to perform the axial exchange with aromatic alcohols due to the fact that the new reactants are not nucleophiles and a lower amount of decomposition side-products are detected. This clean method led us to synthesize three new “exotic” derivatives of singular importance within the chemistry of SubPcs, whose structure has been unambiguously determined by X-ray diffraction.

ASSOCIATED CONTENT Supporting Information

Synthesis, characterization of products and computational de‐

tails. This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Author [email protected]

Tel: +34 91 497 4151. Fax: +34 91 497 3966.

[email protected]

Tel: +34 91 497 2778. Fax: +34 91 497 3966.

ACKNOWLEDGMENT

Financial support from the MINECO, Spain (CTQ-2014-52869-P, T.T.; CTQ2014-57729-P, D.G.-R.), and the Comunidad de Madrid (S2013/MIT-2841 FOTOCARBON, T.T.) is acknowledged. The work of J.G. and L.M.F. was funded by MECD (FPU fellowship).

L.M.F., I.C. and M.Y. thank the MICINN (Spain) for the Project No.CTQ2012-35513-C02-01. The Centro de Computación Científica of the Universidad Autónoma de Madrid is gratefully acknowledged for generous allocation of computational time.

REFERENCES

(1) (a) Claessens, C. G.; González-Rodríguez, D.; Torres, T. Chem.

Rev. 2002, 102, 835; (b) Torres, T. Angew. Chem. Int. Ed. 2006, 45, 2834; (c) Claessens, C. G.; González-Rodríguez, D.; Rodríguez-Mor- gade, M. S.; Medina, A.; Torres, T. Chem. Rev. 2014, 114, 2192.

(2) (a) Morse, G. E.; Helander, M. G.; Maka, J. F.; Lu, Z.-H.;

Bender, T. P. ACS Appl. Mater. Interfaces 2010, 2, 1934−1944. (b) Morse, G. E.; Castrucci, J. S.; Helander, M. G.; Lu, Z.-H.; Bender, T.

P. ACS Appl. Mater. Interfaces 2011, 3, 3538−3544.

(3) (a) Gommans, H.; Aernouts, T.; Verreet, B.; Heremans, P; Me- dina, A.; Claessens, C. G.; Torres, T. Adv. Funct. Mater. 2009, 19, 3435; (b) Beaumont, N.; Cho, S. W.; Sullivan, P.; Newby, D.; Smith, K. E.; Jones, T. S. Adv. Funct. Mater. 2012, 22, 561; (c) Verreet, B.;

Cnops, K.; Cheyns, D.; Heremans, P.; Stesmans, A.; Zango, G.;

Claessens, C. G.; Torres, T., Rand, B. P. Adv. Energy Mater. 2014, 4, 1301413; (d) Cnops, K.; Rand, B. P.; Cheyns, D.; Verreet, B.; Empl, M. A.; Heremans, P. Nat. Commun. 2014, 5, No. 3406.

(4) (a) Sastre, A.; Torres, T.; Díaz-García, M. A.; Agulló-López, F.;

Dhenaut, C.; Brasselet, S.; Ledoux, I.; Zyss, J. J. Am. Chem. Soc. 1996, 118, 746; (b) Martínez-Díaz, M. V.; del Rey, B.; Torres, T.; Agricole, B.; Mingotaud, C.; Cuvillier, N.; Rojo, G.; Agulló-López, F. J. Mat.

Chem. 1999, 9, 1521; (c) de la Torre, G.; Vázquez, P.; Agulló-López, F.; Torres, T. Chem. Rev. 2004, 104, 3723.

(5) (a) Geyer, M.; Plenzig, F.; Rauschnabel, J.; Hanack, M.; del Rey, B.; Sastre, A.; Torres, T. Synthesis 1996, 1139; (b) Potz, R.; Göldner, M.; Hückstädt, H.; Cornelissen, U.; Tutaβ, A.; Homborg, H. Z. Anorg.

Allg. Chem. 2000, 626, 588; (c) González-Rodríguez, D.; Torres, T.

Eur. J. Org. Chem. 2009, 1871; (d) Caballero, E.; Fernández-Ariza, J.;

Lynch, V. M.; Romero-Nieto, C.; Rodríguez-Morgade, M. S.; Sessler, J. L.; Guldi, D. M., Torres, T. Angew. Chem. Int. Ed. 2012, 50, 11337;

(e) Morse, G. E.; Bender, T. P. Inorg. Chem. 2012, 51, 6460.

(6) (a) Tsurumaki, E.; Hayashi, S.; Tham, F. S.; Reed, C. A.; Osuka, A. J. Am. Chem. Soc. 2011, 133, 11956–11959; (b) Tsurumaki, E.;

Sung, J.; Kim, D.; Osuka, A. J. Am. Chem. Soc. 2015, 137, 1056–1059.

(7) Guilleme, J.; Martínez-Fernández, L.; González-Rodríguez, D.;

Corral, I.; Yáñez, M.; Torres, T. J. Am. Chem. Soc. 2014, 136, 14289- 14298.

(8) (a) Xie, Z.; Manning, J.; Reed, R. W.; Mathur, R.; Boyd, P. D.

W.; Benesi, A.; Reed, C. A. J. Am. Chem. Soc. 1996, 118, 2922; (b) Reed, C. A Acc. Chem. Res. 1998, 31, 325; (c) Kato, T.; Tham, F. S.;

Boyd, P. D. W.; Reed, C. A Heteroatom. Chem. 2006, 17, 209.

(9) Guilleme, J.; González-Rodríguez, D.; Torres, T. Angew. Chem Int. Ed. 2011, 50, 3506.

(10) (a) Merling, E.; Lamm, V.; Geib, S. J.; Lacôte, E.; Curran, D.

P. Org. Lett. 2012, 14, 2690; (b) Melen, R. L.; Stephan, D. W. Dalton Trans. 2013, 42, 4795; (c) Müller, M.; Maichle-Mössmer, C.; Bet- tinger, H. F. J. Org. Chem. 2014, 79, 5478.